Taking over an vROPS 8.1 instance and have a few questions about sizing and remote collectors. We currently have one Master and 2 remote collectors. As it is now the remote collectors are installed at same data center as the Master but 9 remote locations point to the 2 remote collectors with few directly to the master node.
If I go to Cluster management on the master node it is showing 57579 Objects in process with only 103 objects being collected, while the remote collectors show 55/3887 objects being collected.
Now if I look at the Collector groups we have default group which the master node resides and another group which the remotes reside. The master node shows number of objects as 341 compared to the managment view of only 103 objects being collected and the remote collectors show 404399/366876 number of objects compared to managment of 55/3887 objects being collected.
I dont quite understand the difference between the two. We have been seeing some performance impacts on the master node both cpu/memory processing aprox 6 million metrics in process on average and have seen constant objects not recieving data from our Horizon adaptors.
Thanks for any help!
make sure you are sizing it correctly. Use the sizing calculator found here https://vropssizer.vmware.com/
That will answer if you are undersized or not.
Horizon adapter missing data means you may have miss some configurations.
you haven't set up the licences correctly
check on the broker to make sure all settings are getting forwarded. All UAG servers and App volumes must be entered in manually for all servers.
This link will help you with the setup.